Calibrating wireless circuitry with digital predistortion
Abstract
An electronic device may include wireless circuitry having a radio-frequency amplifier. A method for calibrating the wireless circuitry is provided that includes characterizing a linear transformation between an input of the radio-frequency amplifier and an input of a feedback receiver, where the radio-frequency amplifier has an output that is coupled to the input of the feedback receiver via a filter and a radio-frequency coupler, obtaining a filter model based on the characterized linear transformation, and generating memory digital predistortion (mDPD) coefficients for a digital predistortion block coupled to the input of the radio-frequency amplifier based on the filter model. The generation of the mDPD coefficients can alternatively be based on an inverse filter model and can further include minimizing an error signal computed from a difference between a predistorted signal output from the digital predistortion block and a demodulated signal output from the feedback receiver.
Claims
exact text as granted — not AI-modifiedWhat is claimed is:
1 . A method of operating wireless circuitry having a radio-frequency amplifier, the method comprising:
characterizing a linear transformation between an input of the radio-frequency amplifier and an input of a feedback receiver, wherein the radio-frequency amplifier has an output that is coupled to the input of the feedback receiver via a filter and a radio-frequency coupler; obtaining a filter model based on the characterized linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver; generating memory digital predistortion (mDPD) coefficients via a learning process while de-embedding the filter model from the learning process; and transmitting a predistorted signal based on the mDPD coefficients.
2 . The method of claim 1 , further comprising:
during the linear transformation characterization, biasing the radio-frequency amplifier in a linear region of operation.
3 . The method of claim 1 , further comprising:
characterizing a linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver at a plurality of frequencies.
4 . The method of claim 1 , wherein the filter model is implemented as a finite impulse response filter or an infinite impulse response filter.
5 . The method of claim 1 , further comprising:
obtaining an inverse filter model based on the filter model, wherein the generation of the mDPD coefficients is based on the inverse filter model and includes filtering, with the inverse filter model, signals output from the radio-frequency amplifier.
6 . The method of claim 1 , further comprising:
applying the mDPD coefficients to a digital predistortion block coupled to the input of the radio-frequency amplifier.
7 . The method of claim 6 , wherein wherein the learning process comprises:
with the filter model, filtering signals output from the digital predistortion block.
8 . The method of claim 6 , wherein the learning process further comprises:
with the digital predistortion block, receiving a baseband signal; with the filter model, filtering the baseband signal to generate a corresponding filtered signal; and with an additional digital predistortion block, receiving the filtered signal and outputting a corresponding signal for generating the mDPD coefficients.
9 . The method of claim 6 , wherein the learning process comprises:
minimizing an error signal computed from a difference between a predistorted signal output from the digital predistortion block and a demodulated signal output from the feedback receiver.
10 . The method of claim 6 , wherein the wireless circuitry is operable in a first power mode and a second power mode different than the first power mode, and wherein the digital predistortion block is deactivated during the first power mode and is activated during the second power mode.
11 . A method comprising:
obtaining a filter model that characterizes effects between an input of a radio-frequency amplifier and an input of a feedback receiver, wherein the radio-frequency amplifier has an output that is coupled to the input of the feedback receiver via a filter and a radio-frequency coupler; obtaining an inverse filter model based on the filter model; generating memory digital predistortion (mDPD) coefficients based on the inverse filter model; and transmitting a predistorted signal based on the mDPD coefficients.
12 . The method of claim 11 , further comprising:
characterizing a linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver while the radio-frequency amplifier is biased in a linear mode.
13 . The method of claim 11 , further comprising:
characterizing a linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver at a plurality of frequencies.
14 . The method of claim 11 , wherein generating the mDPD coefficients comprise:
with the inverse filter model, filtering signals output from the radio-frequency amplifier.
15 . The method of claim 11 , wherein generating the mDPD coefficients comprises:
minimizing an error signal computed from a difference between the predistorted signal and a demodulated signal output from the feedback receiver.
16 . A method comprising:
obtaining a filter model that characterizes effects between an input of a radio-frequency amplifier and an input of a feedback receiver, wherein the radio-frequency amplifier has an output that is coupled to the input of the feedback receiver via a filter and a radio-frequency coupler; generating memory digital predistortion (mDPD) coefficients by filtering, with the filter model, the predistorted digital signal to generate a corresponding filtered signal; and predistorting a digital signal based on the mDPD coefficients.
17 . The method of claim 16 , further comprising:
characterizing a linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver while the radio-frequency amplifier is biased in a linear mode.
18 . The method of claim 16 , further comprising:
characterizing a linear transformation between the input of the radio-frequency amplifier and the input of the feedback receiver at a plurality of frequencies.
19 . The method of claim 16 , wherein generating the mDPD coefficients further comprises:
minimizing an error signal computed from a difference between the filtered signal output from the filter model and a demodulated signal output from the feedback receiver.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.